Skip to content
Sample benchmark demonstrating the VK_NV_cooperative_matrix extension
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

vk_cooperative_matrix_perf is a sample/benchmark demonstrating performance of using the VK_NV_cooperative_matrix Vulkan extension, and the associated GL_NV_cooperative_matrix GLSL extension.

The benchmark queries the supported matrix multiply sizes and precisions from the Vulkan implementation, and runs a couple different shaders at various tiles sizes and reports the performance in teraflops. All matrices are row-major in memory.

Running this application requires an NVIDIA Turing GPU, and a recent driver that supports the VK_NV_cooperative_matrix extension (available from

Modifying and rebuilding the shaders requires a glslangValidator.exe with GL_NV_cooperative_matrix support. This can be built from source using, and will be included in an upcoming SDK release.

You can’t perform that action at this time.